#garfield Logs
Nov 19 2022
#garfield Calendar
12:53 AM rue_mohr: ugh, these baords are bigger, 75c ea
01:11 AM rue_mohr: gettin a list of things on my computer I need to fix
01:11 AM rue_mohr: should I try an SSD again?
01:12 AM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1260816.jpg
01:12 AM rue_mohr: this is a BIG deck
01:12 AM rue_mohr: WHAT is taking up the space?
01:13 AM rue_mohr: there is a bucket of aluminum ingots, that can go to the storage shed if I can set up some shelves
01:13 AM rue_mohr: the relay rack could probably be one of the shevles
01:13 AM rue_mohr: grrr
01:14 AM rue_mohr: there is a truckload of toner cartridges
01:14 AM rue_mohr: I dont even know if the recycling center takes them
01:14 AM rue_mohr: I should find out tommroow
01:15 AM rue_mohr: I think there are still two laser printers to strip down in there
01:15 AM rue_mohr: there are a bunch of big inverters
01:15 AM rue_mohr: 5kw
01:16 AM rue_mohr: 48V
01:16 AM rue_mohr: they wont clean power tho,
01:16 AM rue_mohr: oh I wonder if two of them... no... not enough charger capacity
01:16 AM rue_mohr: I wonder if 4 of them
01:16 AM rue_mohr: 3 charging, and 1 inverting
01:17 AM rue_mohr: oh, but its only 120
01:17 AM rue_mohr: I get 240 out of the genny
01:17 AM rue_mohr: an online ups would be nice
04:34 AM rue_mohr: oooh its 2:30 already
08:04 AM Tom_L: i replaced the supply already and the printer is out the door
03:21 PM rue_mohr: oh I thought it was yours
05:07 PM polprog: damnit, no one except for esera sells eBus adapters
05:07 PM polprog: ebus is a bus for controlling boilers, its RS485 with a DC bias
05:08 PM polprog: but the adapter i have does not pull the bus low enough to send valid symbols
05:09 PM polprog: i wasted 2 months waiting for them to repair it, just to install it and see the same fault on the scope
05:09 PM polprog: i will see what their reply is
05:09 PM polprog: and if they dont provide a solution i will keep digging, check the bus currents and so on
05:45 PM rue_mohr: *transistors* just sayyin
05:51 PM rue_mohr: so, 10 and 20V
05:52 PM rue_mohr: no reverse tho eh? just 10/20V
05:55 PM rue_mohr: polprog, you still there?
05:56 PM rue_mohr: your not are ya
05:56 PM rue_mohr: you just surfaced to drop that and dissapear again
05:58 PM rue_mohr: I wonder if someone didn't use a current limited supply on your ebus
05:59 PM rue_mohr: should be limited to 50mA
05:59 PM rue_mohr: and in no cases over
06:00 PM rue_mohr: 100mA
06:00 PM polprog: yea
06:00 PM rue_mohr: I should just make a wall to talk to
06:00 PM rue_mohr: oh you are here
06:00 PM polprog: im here ;)
06:00 PM rue_mohr: so, do you just need a level convter?
06:00 PM rue_mohr: like, is it equipment or a computer or what are you up to?
06:01 PM rue_mohr: ebus-wiki.org/lib/exe/fetch.php/ebus/spec_prot_12_v1_3_1_e.pdf
06:01 PM rue_mohr: there is a full set of schematics in that doc
06:01 PM rue_mohr: eh?
06:01 PM rue_mohr: :( I wanted to help design one
06:01 PM polprog: its a boiler
06:01 PM polprog: https://polprog.net/rozne1/ircjunk/ebus/img/mainboard.jpg
06:02 PM rue_mohr: so, someone MADE an ebus chip that its communicating with?
06:02 PM rue_mohr: or you replaced the board?
06:03 PM polprog: Someone made a box that is a usb to ebus converter
06:03 PM polprog: https://polprog.net/rozne1/ircjunk/ebus/img/17a_small22.jpg
06:03 PM polprog: and im trying to make it talk to the board
06:03 PM polprog: the board is OK
06:03 PM polprog: it runs the boiler
06:03 PM rue_mohr: ok, so who is supplying power to the bus?
06:03 PM polprog: but the converter puts out undefined voltages, not pulling low enough
06:03 PM polprog: The mainboard is supplying power
06:03 PM rue_mohr: oh, why not make the converter in that doc?
06:04 PM polprog: i dont have time for it really
06:04 PM rue_mohr: well
06:04 PM polprog: I will probably eventually
06:04 PM polprog: but i was hoping to have a working solution
06:04 PM rue_mohr: if you take a meter, whats the short circuit current on the ebus?
06:04 PM rue_mohr: maybe its over spec
06:04 PM polprog: i will have to check it
06:04 PM polprog: thats a possibility
06:05 PM rue_mohr: they have everything in here
06:05 PM rue_mohr: they transmit by grounding to a 7.5V zener
06:06 PM polprog: https://polprog.net/rozne1/ircjunk/ebus/ebus.html
06:09 PM rue_mohr: heh
06:09 PM rue_mohr: did you write an append system for blogging ;]
06:09 PM polprog: no, i wrote all of this today in markdown then just
06:10 PM rue_mohr: ah
06:10 PM rue_mohr: so
06:10 PM rue_mohr: you dont have time to make an adatper
06:10 PM rue_mohr: if I have time, I should make one
06:10 PM rue_mohr: and send it to you
06:10 PM polprog: ~ircjunk/ebus$ markdown_py <ebusd.md >ebusd.html
06:10 PM rue_mohr: it would take about a month
06:10 PM polprog: hmm
06:10 PM rue_mohr: +- the time for me to design it
06:10 PM polprog: i need it before the winter is over ;)
06:10 PM rue_mohr: winter lasts 8 months
06:11 PM rue_mohr: oh, well, here in canada it does
06:11 PM polprog: 2 months here
06:11 PM rue_mohr: hmm
06:11 PM polprog: i think i will validate the bus current, then take the green box apart and see what the problem is
06:11 PM polprog: how hard is to pull down A to B with a get
06:11 PM rue_mohr: how long does it take osh to get a pcb to you?
06:11 PM polprog: fet*
06:11 PM polprog: i never ordered osh, but takes long time
06:11 PM rue_mohr: that protocol is super easy
06:11 PM rue_mohr: hmm
06:12 PM rue_mohr: how fast can you otherwise get a kicad pcb made?
06:12 PM polprog: 2 week
06:12 PM rue_mohr: that would leave one week with heat at the end of winter
06:13 PM rue_mohr: well, maybe you should pay someone too much for a *now* solution?
06:14 PM polprog: there is an open solution, the problem is they only make a batch when people sign up
06:14 PM polprog: it also needs to be compatible with ebusd
06:14 PM rue_mohr: all the schematics are in that pdf
06:15 PM polprog: yes and no, ebusd does some arbitration and it can be done in hardware
06:15 PM rue_mohr: just stick on a ttl->usb
06:15 PM polprog: It's a goddamn docu nightmare
06:15 PM rue_mohr: but your adapter is just serial
06:15 PM rue_mohr: its just 1200 baud serial
06:15 PM polprog: I havent taken it apart yet, its a commercial one
06:15 PM rue_mohr: the rest is worked out in software
06:15 PM polprog: Well no, because some adapters do arbitration in hardware
06:15 PM polprog: thats why i say its a nightmare
06:16 PM rue_mohr: uh
06:16 PM polprog: it looks like usb->ttl, turns out to be usb->rs485 with weird voltages
06:16 PM polprog: and even then the B wire is held at constant +12 IIRC, unlike RS485 where its driven
06:16 PM polprog: i can waste weeks here
06:16 PM polprog: and i dont want to do that
06:16 PM rue_mohr: no its just a ground wire and a +v that fluctuatates
06:17 PM rue_mohr: did you open that pdf?
06:17 PM polprog: i did
06:17 PM polprog: ah
06:17 PM polprog: yeah its at ground
06:17 PM polprog: i was poking it with a scope and it was flat
06:18 PM rue_mohr: some things are a lot simpler than they seem...
06:18 PM rue_mohr: I woke up just as the sun finished setting
06:18 PM rue_mohr: I hate winter
06:18 PM polprog: :(
06:20 PM rue_mohr: I want to work on a project with someone
06:20 PM polprog: same :(
06:20 PM rue_mohr: nobody wants to work togethor on anything tho
06:20 PM polprog: i will hopefully have more time when the thesis is done
06:20 PM polprog: im close
06:20 PM polprog: i want to!
06:20 PM rue_mohr: and you have no time for anything
06:20 PM rue_mohr: EVER
06:20 PM polprog: well, now yes..
06:20 PM rue_mohr: pff
06:20 PM polprog: i've been working hard for the last 3 years
06:21 PM polprog: remember the minimum parts count oscillators?
06:21 PM polprog: we used to sit down and build stuff over irc for a few hours
06:21 PM rue_mohr: :] we did good
06:21 PM rue_mohr: I didn't think an SCR oscillator could be done
06:21 PM polprog: :) you did one
06:22 PM polprog: i did a different one
06:22 PM polprog: we did well :)
06:24 PM rue_mohr: I have 12 pcbs comming from osh
06:25 PM rue_mohr: total $2.50
06:25 PM polprog: wow
06:25 PM polprog: yeah what kills osh for me is the shipping
06:25 PM rue_mohr: minimum circuits pay off!
06:25 PM polprog: wrong continent
06:25 PM rue_mohr: awe
06:25 PM rue_mohr: I think there are similar and better services over there tho
06:26 PM rue_mohr: did you see the link to joker?
06:26 PM polprog: no?
06:26 PM rue_mohr: they just throw your design on whatever board is available?
06:26 PM polprog: ah that
06:26 PM polprog: yeah, the price is ajoke ;)
06:26 PM rue_mohr: I couldn't quite read it, but the prices looked high
06:26 PM rue_mohr: no silkscreen either
06:26 PM polprog: they were extremely high
06:26 PM rue_mohr: ok
06:26 PM polprog: typical germany
06:27 PM rue_mohr: who do you go to for pcbs?
06:27 PM polprog: thats like the polish pcb factory that quoted me $200+ for a "prototype" run, the price included negative storage for a year
06:28 PM polprog: i usually do JLC, or some local company which does a group buy in JLC or some other chinese house
06:28 PM polprog: cause the labels are the same as in JLC, they repack to their own boxes
06:30 PM rue_mohr: zippo:/files/programming/src/polprog# gcc lambda.c
06:30 PM rue_mohr: zippo:/files/programming/src/polprog#
06:30 PM rue_mohr: hahah, it ate it with no complaints
06:31 PM polprog: ;)
06:32 PM rue_mohr: whats important is that now I can close a browser tab
06:32 PM polprog: haha
06:33 PM rue_mohr: only like 130 tabs left
06:33 PM rue_mohr: across ... 13 windows, across 3 users
06:35 PM polprog: 3 users?
06:37 PM rue_mohr: funny things happen if you run all the browsers as the same user
06:37 PM rue_mohr: so I broke them up
06:37 PM rue_mohr: less crashing
06:38 PM rue_mohr: I'm excited for my pcbs to arrive
06:38 PM rue_mohr: I did some fet drivers and some oscillators
06:38 PM polprog: ahh
06:38 PM rue_mohr: the fet driver is a DIP4, the oscillators are a SIP4
06:38 PM polprog: I need to clean my desk and build the power supplt
06:38 PM polprog: i ordered new solder paste
06:39 PM rue_mohr: I know lots more about smps now, I should re-attack my big power supply
06:39 PM rue_mohr: I'm still waffling on a final analog stage
06:40 PM rue_mohr: and have the digital track the analog+2V
06:41 PM rue_mohr: 5A
06:41 PM polprog: hmmm
06:41 PM polprog: yeah i was considering that
06:41 PM rue_mohr: so, probably between 1 to 10w loss out the analog stage
06:41 PM rue_mohr: pretty good for 30A 5A
06:41 PM polprog: i need a clean analog supply -24V for the microwave gen
06:42 PM polprog: tuning V
06:42 PM polprog: i have it now made by a buck, with litle load and large caps it should be clean
06:42 PM rue_mohr: yup
06:42 PM rue_mohr: I seem to be obsessed with power supplies right now
06:42 PM polprog: but i was considering making a (-24-x) V buck that feeds into a negative linear reg witha drop of x
06:43 PM rue_mohr: I'm trying to suppress it, but I'm eager to make something
06:43 PM polprog: or i could even run the reg's output to the bucks feedback to make it self regulating
06:43 PM rue_mohr: you only need maybe 0.5mA eh?
06:44 PM polprog: er, i may need more
06:45 PM polprog: it will be powering the entire rack
06:45 PM rue_mohr: did you post that one?
06:45 PM polprog: but no more than say 100mA, it will always be the varicap voltage
06:45 PM polprog: yes, ages ago
06:45 PM rue_mohr: I really got an intersting feel for current consumption with the lower power oscillator thing
06:45 PM polprog: https://twitter.com/polprogpl/status/1567970067622633472
06:46 PM rue_mohr: 20mA is great
06:46 PM rue_mohr: for a supply
06:46 PM polprog: its a tangent from the minirack build
06:46 PM polprog: anyway
06:46 PM rue_mohr: a good one can run on about 6mA
06:46 PM polprog: i have to run to seel
06:46 PM polprog: sleep*
06:46 PM rue_mohr: and then the cd40106 at freaking 76uA
06:46 PM polprog: i gotta go to the bazar to look for unusual junk tomorrow at 8
06:46 PM rue_mohr: hah, facetime my cellphone!
06:46 PM rue_mohr: :] oh well
06:46 PM polprog: haha
06:47 PM polprog: i will post pics to ircjunk :)
06:47 PM rue_mohr: thanks
06:47 PM polprog: night!
06:47 PM polprog: and thanks :)
06:47 PM rue_mohr: we nee to work out dynamic blogs
06:47 PM polprog: Markdown is good, just add date, append, rebuild
06:47 PM rue_mohr: I dont like having to compile
06:48 PM rue_mohr: BUT
06:48 PM polprog: make a cron job
06:48 PM rue_mohr: I have to run a script to post anyhow
06:48 PM polprog: ;)
06:48 PM polprog: its easy
06:48 PM polprog: allright, see you on the flip side
06:48 PM rue_mohr: oh, gnight
06:48 PM rue_mohr: see you in the morning
06:48 PM rue_mohr: 8 hours?
06:48 PM polprog: zzz
06:48 PM polprog: yea
06:48 PM rue_mohr: midnight, see you there
09:39 PM rue_mohr: http://www.linusakesson.net/hardware/chiptune.php
09:39 PM Tom_L: you know buildbot?
09:39 PM Tom_L: or git
09:40 PM Tom_L: stuck on one command
10:00 PM rue_mohr: nope
10:00 PM rue_mohr: I mean, try me, but
10:03 PM Tom_dev: http://docs.buildbot.net/current/manual/configuration/steps/gitcommit.html
10:04 PM Tom_dev: bf.addStep(steps.GitCommit(workdir='build', messages=['commiting deb'],
10:04 PM Tom_dev: paths=['build/howdy_0.0.1_amd64.deb'], emptyCommits='disallow', logEnviron=True))
10:04 PM Tom_L: is the line i'm working on
10:04 PM Tom_L: can't seem to get the syntax quite right
10:05 PM rue_mohr: python?
10:06 PM Tom_L: buildbot and yes it's python
10:06 PM Tom_L: 3
10:06 PM rue_mohr: are there any errors to go by?
10:06 PM Tom_L: i'll get you some shortly
10:14 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] releaseLocks(GitCommit(workdir='build', messages=['commiting deb'], paths=['build/howdy_0.0.1_amd64.deb'], emptyCommits='disallow', logEnviron=True)): []
10:14 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] step 'gitcommit' complete: failure (['finished (failure)'])
10:14 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] <Build buildtests number:39 results:failure>: build finished
10:14 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] releaseLocks(GitCommit(workdir='build', messages=['commiting deb'], paths=['build/howdy_0.0.1_amd64.deb'], emptyCommits='disallow', logEnviron=True)): []
10:14 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] step 'gitcommit' complete: failure (['finished (failure)'])
10:14 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] <Build runtests number:38 results:failure>: build finished
10:14 PM rue_mohr: I'm going to randomly guess that the user its running as does not have access to do the commit
10:15 PM Tom_L: should, keys are in place
10:15 PM Tom_L: but the commit should be local
10:15 PM Tom_L: the push will be to the repo
10:16 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] releaseLocks(ShellCommand(workdir='build/bbtest', command=['debuild', '-us', '-uc'])): []
10:16 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] step 'shell' complete: success (None)
10:17 PM Tom_dev: that is the last command that is successful
10:17 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] <RemoteShellCommand '['git', 'add', 'build/howdy_0.0.1_amd64.deb']'>: RemoteCommand.run [23]
10:17 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] command '['git', 'add', 'build/howdy_0.0.1_amd64.deb']' in dir 'build'
10:17 PM Tom_dev: 2022-11-19 22:09:43-0600 [-] Source step failed while running command <RemoteShellCommand '['git', 'add', 'build/howdy_0.0.1_amd64.deb']'>
10:18 PM Tom_dev: that is one before the commit
10:18 PM Tom_dev: but if i comment out the commit line i get build success on all
10:19 PM Tom_dev: it's hard to follow the messages because they seem to repeat
10:19 PM rue_mohr: build is compiling
10:19 PM rue_mohr: it'll compile in either case\
10:19 PM Tom_dev: yes, i get a deb
10:19 PM rue_mohr: but if it cant send it, it says failure
10:20 PM Tom_dev: i don't know if it's a path issue, name issue etc
10:20 PM rue_mohr: I think you need to verify if It can commit to git
10:20 PM rue_mohr: screw the compiling
10:20 PM Tom_dev: i've done others using git gui just fine
10:21 PM Tom_dev: it has a key listed for this pc
10:21 PM rue_mohr: your not using the gui tho
10:21 PM Tom_dev: no
10:21 PM rue_mohr: your using a script that running as a user that has different permissions
10:21 PM Tom_dev: but the key should be enough i think
10:21 PM rue_mohr: make sure of what user its saying it is
10:21 PM rue_mohr: lots of scripts change to a different user for security
10:21 PM Tom_dev: it's coming from a bot
10:22 PM rue_mohr: yea
10:22 PM rue_mohr: find out who its running as
10:22 PM rue_mohr: username
10:22 PM Tom_dev: messenger
10:22 PM rue_mohr: if you can, get it to print the username to a text file
10:22 PM rue_mohr: then sudo a test command as that user
10:22 PM rue_mohr: sudo -u messenger git commit...
10:32 PM Tom_dev: running in a vm
10:32 PM Tom_dev: can't
10:33 PM rue_mohr: you can make a step that does it for you (to get the user name)
10:33 PM Tom_dev: said user not found
10:33 PM Tom_dev: unknown user
10:33 PM rue_mohr: whoami ?
10:33 PM rue_mohr: make a step that runs whoami and sends it to a text file
10:33 PM rue_mohr: or prints it or something
10:39 PM Tom_dev: 'SSH_AUTH_SOCK': '/run/user/1000/keyring/ssh',
10:39 PM Tom_dev: 'TERM': 'xterm-256color',
10:39 PM Tom_dev: 'USER': 'tom',
10:40 PM Tom_dev: that part is right
10:40 PM Tom_dev: i think it's the path to the file
10:40 PM Tom_dev: it actually fails on the add not the commit. both are a part of the gitCommit command in buildbot
11:14 PM rue_shop3: I cant focus on anything
11:15 PM Tom_L: i give up for tonight
11:16 PM Tom_L: a) don't know python
11:16 PM Tom_L: b) don't know buildbot
11:16 PM Tom_L: c) don't know git that well
11:17 PM Tom_L: but at least it does build
11:17 PM Tom_L: commit a change on github and the bot is triggered for a local build and reports it to irc
11:18 PM Tom_L: i just can't get the deb pushed back to github
11:32 PM rue_mohr: k